Can Terrorism Ever Be Morally Justified?

Abstract

This paper provides a framework to make moral sense of terrorism. The framework consists in a test, referred to as the MODAL test, which is an acronym standing for five tests or principles for determining the moral defensibility or indefensibility of terrorism. The five principles concern the motives for terrorism, its objectives, destructiveness, availability of alternatives, and likelihood of success. This approach makes it conceivable but highly unlikely in practice that a terrorist act is morally justified. The MODAL test does not claim to be an exhaustive framework for analysing the moral legitimacy or illegitimacy of terrorism but rather a practical analytical tool aimed at securing a reliable grasp of the tricky question of the relation between morality and terrorism.

I

In a ‘letter to the America people’ written in the aftermath of the 9/11 attacks, Osama bin Laden set out to answer two questions: why are we fighting you, and what do we want from you?¹ His answer to the first question is simple: ‘because you attack us and continue to attack us’. After asserting that ‘you attacked us in Palestine’, he adds that ‘the blood pouring out of Palestine must be avenged’ and that ‘it is commanded by our religion and our intellect that the oppressed have a right to return the aggression’. Given that the Almighty has ‘legislated the permission and the option to take revenge’, bin Laden offers the American people this piece of advice: ‘Do not wait for anything from us but Jihad, resistance, and revenge’. For ‘if we are attacked, then we have the right to attack back. Whoever has destroyed our villages and towns, then we have the right to destroy their villages and towns. Whoever has stolen our wealth, then we have the right to destroy their economy. And whoever has killed our civilians, then we have a right to kill theirs’.

The answer that bin Laden gives to his second question is multi-faceted. He describes America as ‘the worst civilization witnessed by the history of mankind’ and calls on it to turn to Islam. He advises America to stop supporting Israel and to ‘pack your luggage and get out of our lands’ so as not to ‘force us to send you back as cargo in coffins’. In his letter, bin Laden also tries to justify the targeting of civilians on 9/11 ‘for crimes they did not commit and offences in which they did not partake’.² His justification is that American civilians, including the occupants of the World Trade Center, were not innocent since they have ‘chosen, consented to, and affirmed their support for the Israeli oppression of the Palestinians’.³ He takes no account of the deaths of foreigners on 9/11 or of Americans victims of terrorism who did not support the policies of their government.

The letter makes it clear that bin Laden saw the need to demonstrate that al-Qaeda’s terrorism on 9/11 was morally justified. Most terrorists are convinced that their actions are morally justified but few provide such detailed justifications as the ones provided by bin Laden. His arguments might be dismissed on the grounds that terrorism is morally wrong by definition and can never be morally justified because it involves the murder of innocents. There are several reasons for not proceeding in this way. As the letter shows, there is some dispute about who counts as ‘innocent’. One might take the view that all civilians are innocents, contrary to what bin Laden argues, but terrorism does not necessarily harm civilians. The al-Qaeda attack on the USS Cole in Aden was a terrorist attack on a military target.⁴

In any case, as Virginia Held argues, the question whether terrorism can ever be morally justified ‘cannot be answered by resort to definitional fiat’.⁵ Readers may not need much convincing that bin Laden was wrong about the morality of 9/11, but other examples are less straightforward. When the French Resistance planted bombs in public places as part of its campaign against Nazi occupation, its actions were terroristic in character and some former members have described themselves as terrorists.⁶ The philosopher R. M. Hare gives this as an example of terrorism that could be morally justified.⁷ Hare might be wrong about this, but he is not obviously wrong. There is room for debate about whether terrorism can ever be morally justified. In remarks on a brutal terrorist attack on Israel by Hamas in 2023, President Biden insisted that ‘there is no justification for terrorism. There is no excuse’.⁸ His outrage about this particular terrorist act was understandable but are we prepared to say that there was no excuse for terrorism in the context of the Nazi occupation of France?

In seeking to provide a moral justification for the 9/11 attacks, bin Laden does two things. First, he identifies the aims or objectives of the attacks. Setting aside the religious aims, the primary political objective was to induce America to change its actions and policies in the Middle East. The moral worthiness (in his eyes) of this objective was plainly regarded by bin Laden as contributing to the moral justification of 9/11. The underlying objectives principle is that:

(O) Terrorism is morally justified only if its objectives are morally justified or morally worthy.

In addition to describing the objectives of al-Qaeda terrorism, bin Laden outlines his motives or motivating reasons for ordering the 9/11 attacks. The motivating reasons for an action are the considerations in the light of which the agent acted and which he took to favour acting as he did. Thus, a second principle might be:

(M) An act of terrorism is morally justified only if has morally sound or at least morally acceptable motives.

This is the motives principle. A key terrorist motive is revenge or retribution. Indeed, for many terrorists, retribution is both a motive and a key terrorist objective. In his analysis, J. Angelo Corlett defines retributive terrorism as ‘terrorism that aims at giving a person or group what they “had coming to them,” e.g., giving them what they deserve’.⁹ There is no doubt that in bin Laden’s eyes the USA had it coming, and that he saw this as a moral consideration in favour of the 9/11 attacks.

The objectives principle is impossible to apply unless it is clear what counts as a morally worthy objective. There will always be disagreements about the morality of terrorism because there will always be disagreements about what constitutes a ‘morally worthy’ objective. In the same way, there will always be disagreements about what counts as a morally sound motive and, in particular, about the moral standing of the motive of revenge. In this case, there is the additional complication that for some philosophers, such as J. S. Mill, the moral rightness of an action does not depend on the agent’s motives.¹⁰ A contrary view is that ‘an act is morally acceptable if an only if it comes from good or virtuous motivation’.¹¹ I will have more to say about (O) and (M) below.

To keep things simple at least for the moment, let us assume that the moral rightness of an action is not wholly independent of the agent’s motives and that we have at least an intuitive grasp of what counts as a morally sound motive and a morally worthy objective. Are we then in a position to say that an act of terrorism is morally justified if and only if both its objectives and underlying motives are morally sound? Not at all, since there are at least three other issues that need to be taken into consideration: the likelihood of success, the availability of alternative, non-terroristic means of achieving the same result, and the nature of the terrorism used to attain a given objective. Terrorism can be more or less destructive, and a natural thought is that only a minimally destructive form of terrorism has any chance of being morally justified.

Even if a given action has both a morally worthy objective and a morally sound motive, it will still not be morally justified if there is little prospect that it will deliver the desired result. The chances of the 9/11 attacks inducing the United States to withdraw from the Middle East were slim, and so it proved. Indeed, 9/11 had the effect of increasing the American footprint in the Middle East. The attacks would have been objectionable from a moral point of view even if they had turned out to be politically effective. However, the fact that they were not effective, and were never likely to be, made them more objectionable. The broader question is whether terrorism works.¹² If it does not deliver either tactical or strategic success, or only does so rarely, this is not just a practical objection to terrorism. It is also a moral objection since it adds to the sense that terrorism wastes human lives. Thus, even if a given act of terrorism has a morally worthy objective, a further principle is:

(L) Terrorism is morally justified only if it is likely to achieve its objective.

A question raised by this likelihood principle is what to make of terrorists acts whose objective is revenge. Louise Richardson implies that terrorism cannot fail to achieve this objective since ‘terrorists need rely only on themselves to get revenge. They take it; it is not given to them’.¹³ I will return to this claim below. Meanwhile, suppose that an act of terrorism is likely to attain a morally worthy objective and has a morally sound motive. Even then, the act is still objectionable if there are equally effective, alternative means of attaining the same end, that is, alternative means that are non-terroristic. Thus, according to what I call the alternatives principle:

(A) Terrorism is morally justified only if there are no realistic non-terroristic means of achieving its objective.

The intuition underpinning this principle is that for terrorism to be morally justified it must be necessary.

Even if there is no effective alternative to terrorism in some form, there is terrorism and there is terrorism. The 9/11 attacks were designed to be maximally destructive, that is, to kill as many people as possible and to cause as much physical damage as possible. Their conception and design also ensured that most victims would be civilians who were not responsible for the policies to which bin Laden objected. The Hamas attack on Israel in October 2023 was not only maximally destructive but also maximally sadistic. The rape, torture, and mutilation of victims amounted to what Adriana Cavarero calls horrorism.¹⁴ Horrorists are ‘not content to kill because killing would be too little’.¹⁵ Their sadism and brutality serve no obvious strategic purpose. Rather, the use of extreme methods is simply designed to maximize suffering. The evident glee with which Hamas terrorists slaughtered their victims made their actions all the more morally repugnant.

Suppose that such extreme terrorism is described as maximal terrorism. A question that might be raised at this point is whether maximal terrorism would be morally justified if it were the only possible means of attaining the terrorists’ objectives. If genocide is the objective, then only genocidal methods will do. However, since there are no circumstances in which genocide would be a morally acceptable objective, this concern can be set aside. I can think of no realistic circumstances in which a morally worthy objective could only be achieved by maximal terrorism.¹⁶ For terrorism to be morally justified, it must be minimal rather than maximal: it must minimize the death and suffering of its victims, especially innocent victims, and the destruction of property. This constraint is captured by the following destructiveness principle:

(D) An act of terrorism is morally justified only if it is a minimally destructive means of achieving its objective.

A rationale for this principle, which presupposes a morally worthy objective, is that the alleged benefits of terrorism are more likely to outweigh the harms it causes if harms are minimized. The harms cause by maximal terrorism are so great and so egregious as to make it impossible in practice for them to be outweighed by any supposed benefits.

I now have a framework for addressing the question whether terrorism can ever be morally justified. Suppose that X is an action that qualifies as an act of terrorism according to my definition of terrorism. Whether X is morally justified depends on the agent’s motives and objectives, the destructiveness of X, the availability of alternative means, and the likelihood of success. I call these five tests the MODAL test. The fact that an act of terrorism passes this test entails that it is morally justified only if the MODAL conditions are not only individually necessary conditions for moral justification but also jointly sufficient. I make no claim of joint sufficiency, and do not rule out that possibility that an act that passes the MODAL test is still not morally justified because it fails to satisfy further necessary conditions or because there are additional factors that defeat the claim of moral justification. Philosophers almost invariably run into trouble when they claim to have discovered sufficient conditions for something to be the case; counterexamples are always in the offing.

A more realistic approach, and my approach here, is to address those who argue that terrorism cannot be morally justified because one or more putative necessary conditions for moral justification cannot be satisfied. In my view, MODAL identifies five plausible necessary conditions for terrorism to be morally justified and it is possible, though difficult, for an act of terrorism to satisfy these conditions. To insist that no terrorist act can pass the test is to betray a lack of imagination about the circumstances in which people can find themselves. It is one thing to allow that an act of terrorism that passes the MODAL test is still not morally justified. However, compelling reasons will need to be given for supposing that a given terrorist act is a case of this type. For example, suppose that an act of terrorism passes the MODAL test, but its unintended long-term consequences are disastrous. This would be a proper reason for questioning whether the act was morally justified despite passing MODAL. In the absence of such reasons, it is a reasonable—albeit defeasible—hypothesis that the act is indeed morally justified in virtue of satisfying the five MODAL conditions.

When philosophers suggest that some actions of the French Resistance are examples of morally justified acts of terrorism, we interpret this as the claim that they pass the MODAL test or at least satisfy the five MODAL conditions to some degree. The idea that passing the MODAL test is not all or nothing and that acts can pass or fail more or less comprehensively is an important one, although I will not make much of it in this paper. An implication of this approach is that moral justification is not all or nothing. It makes sense to think in terms of degrees of moral justification and to suppose that one act of terrorism is more justified from a moral point of view than another. Many acts of terrorism, such as 9/11 and the Hamas attack on Israel in October 2023, comprehensively fail the MODAL test. It may be that no terrorist act passes the test as comprehensively as many terrorist acts fail it.

The idea that acts of terrorism can at least in principle be morally justified should be less troubling for sceptical readers once they see that it is no easy matter to pass the MODAL test. I think we can be reasonably confident that those who insist that terrorism can never be morally justified would revise their views in certain cases, for example, ones in which they have to choose between resorting to terrorism and acquiescing in their own annihilation or subjugation. For example, if it had been possible for an act of terrorism or, more realistically, a terrorist campaign to prevent the Holocaust, it would be counterintuitive to insist that terrorism would not have been justified in these circumstances. In Michael Walzer’s terminology, the Holocaust represented a supreme emergency for Jews in Germany and elsewhere.¹⁷ It might be objected that this is a special case and that we should not infer from the fact that terrorism might be morally justified in a supreme emergency that it can be morally justified in other circumstances. Terrorists always think that they are facing a supreme emergency but what matters is whether their belief is correct. I see the force of this line of reasoning but am not persuaded that terrorism can only be morally justified in a supreme emergency on the scale of the Holocaust. The MODAL test is demanding but not that demanding. The next challenge, therefore, is to put some bones on the test, starting with the objectives principle. I will postpone discussion of (M) until the end since it is the most contentious MODAL principle.

II

One and the same act can have multiple different objectives, some of which are tactical while others are strategic. There are also the personal objectives of the individuals responsible for carrying out an act of terrorism and the personal or political objectives of those (if different) who ordered the attack. Consider the case of Mohammad Atta, who flew American 11 into the North Tower of the World Trade Center on 9/11. His immediate objective was to destroy the building and kill as many people as possible. His other objectives perhaps included martyrdom and renown. However, bin Laden’s objectives in ordering the 9/11 attacks were not the same. A tactical objective was reaction: he hoped to provoke America to overreact in a way that would increase support for al-Qaeda in the Muslim world.¹⁸ His overarching strategic objective was to induce a major change in American foreign policy and at the same time to punish ‘the worst civilization witnessed by the history of mankind’ for its past sins. These were not so much the objectives of a single act of terrorism but of a wider terrorist campaign that included the 7/7 bombings in London and mass casualty attacks in Bali, Madrid, and East Africa.

The objectives principle says that terrorism is morally justified only if its objectives are morally justified or morally worthy but whose objectives and which objectives are in question in this formulation? Louise Richardson distinguishes short-term organizational objectives from ‘long-term objectives requiring significant political change’.¹⁹ The former, which in reality are as likely to be personal as organizational objectives, are the three Rs: revenge, renown, and reaction. The moral status of revenge will be considered below. Personal renown is hardly a morally worthy objective and the same goes for reaction if considered in isolation. The moral worth of reaction and the moral justification for mass killing derives, at least in the eyes of the terrorist, from the moral worth of the strategic political objectives to the achievement of which they are taken to make a significant contribution. Thus, when (O) refers to the moral worth of the objectives of terrorism, the latter should be understood as its long-term strategic objectives. The moral worth of terrorism, if any, derives from the moral worth of the political change or changes that are its ultimate objective.

What kinds of significant political change do terrorists seek, and what could their moral worth or moral value possibly consist in? Virginia Held observes that:

If the violence of terrorism is used for a purpose that is morally wrong, such as forcibly to impose a given religion on a group of persons and to kill all who resist, such terrorism can rather obviously be seen to be morally wrong. More frequent and difficult cases are whether terrorism can justifiably be used for morally admirable objectives, such as to liberate groups from colonial or racial oppression.²⁰

This framing is very much in line with the objectives principle. Terrorism is morally wrong if it is used for a purpose that is morally wrong. It is not necessarily morally wrong if it is used for morally admirable objectives. These include liberation from colonial or racial oppression. Liberation from other forms of severe oppression might also qualify. Held has written in this connection about societies in which the rights of large numbers of people are denied. Terrorism violates the rights of its victims, but she invites us to weigh this fact against the possibility that it protects or promotes the rights of the majority. Consider the struggle of the African National Congress (ANC) against apartheid in South Africa. Since apartheid involved the grossly unjust violation of the rights of the majority, Held infers that ‘if the ANC had engaged in substantial terrorism, an impartial judgement would likely conclude that the violence used by the apartheid government in attempting to suppress it was more unjustified than the violence used by the ANC’.²¹

In fact, some of the ANC’s actions were undoubtedly acts of terrorism, even if terrorism played a relatively minor role in its struggle. In campaigning against apartheid, the ANC also campaigned for democracy. Suppose that a transition to democracy was one of its strategic aims and that terrorism contributed to the achievement of this aim. This would make its terrorism a form of what Ted Honderich calls ‘democratic terrorism’, terrorism whose intended and perhaps even actual result is democracy.²² This sounds paradoxical. How can terrorism ‘be directed to undeniably good ends, first of all democracy’ given that it ‘consists, in part, in atrocity and carnage’?²³ There is no doubt that means and ends can conflict, and that terrorism can be, and often is, destructive of democracy. However, the example of the ANC suggests that terrorism is not necessarily destructive of democracy and can even play a role in facilitating the transition from a non-democratic to a democratic government.²⁴ South African democracy has its problems, but these are not the result of the occasional use of terrorism in the ANC’s fight for democracy.

Democracy and liberation from oppression are among the morally good ends to which terrorism can be and has been directed. They were among the ends of the French Resistance, the ANC, and some anti-colonial campaigns. However, Honderich’s claim that democracy is an undeniably good end might be questioned since its goodness has been disputed by Islamist thinkers who regard it as incompatible with sovereignty of God. This raises a difficult question: given that most terrorists see their objectives as morally worthy, how are we to distinguish between ends that are actually morally worthy and ones that are not? Some might object to the implication that there is a fact of the matter about moral worthiness. I take it that democracy and respect for human rights are genuinely worthy objectives from a moral point of view but what are we to say to those with a radically different conception of what counts as a morally worthy or admirable objective of political action?

This is too large a question to be satisfactorily answered here but the Aristotelian moral framework developed by the philosopher Philippa Foot gives at least an indication of what a satisfactory response might look like. She notes that ‘for all the diversities of human life, it is possible to give some quite general account of human necessities, that is, of what is quite generally needed for human good’.²⁵ Freedom from oppression, respect for human rights, and, arguably, democracy are among the objective conditions for human flourishing or what Foot calls ‘human good’. This accounts for their worthiness as political objectives. We do not expect universal acceptance of the liberal democratic view of human flourishing, but our remarks are primarily addressed to those who do accept it. Assuming that democracy, freedom, and respect for human rights are morally worthy objectives and can be among the objectives of terrorism, it follows logically that terrorism can have morally worthy objectives.

To be clear, my claim is not that terrorism is morally justified if it has morally worthy objectives. It is that terrorism is morally justified only if it has morally worthy objectives. (O) only states a necessary condition for terrorism to be morally justified rather than a sufficient or necessary and sufficient condition. However, if we think that democracy, freedom, and respect for human rights matter then we must also accept that they might be worth fighting for. Indeed, these are precisely the values for which many countries did fight in the Second World War. If states can legitimately use violence in defence of democratic values, then it is not clear why sub-state actors should not be permitted in some circumstances to resort to violence in pursuit of the very same objectives. The politically motivated violence of sub-state actors is more likely to be called ‘terrorism’ than that of state actors but why should this make a difference from a moral point of view?

Plainly, this discussion of the objectives principle leaves many questions unanswered. For example, liberation from oppression might be a morally worthy objective but how bad does oppression need to be for it to contribute to a moral justification of liberatory terrorism, as it might be called? The assumption is that terrorism is only a potentially justifiable response to severe oppression, such as that experienced by Jews in Nazi Germany. A person might be persuaded that democracy is worth fighting for but is it important enough to contribute to the justification of terrorism? Much depends on the alternative to democracy. In South Africa, the alternative was apartheid. Against that background, democratic terrorism is much easier to justify than if the status quo is undemocratic but the democratic deficit is a small one. Given what terrorism involves, it cannot be morally justified by mild repression or minor departures from democratic ideals.

Given that most terrorists in the world today are fighting for objectives that are very far from morally admirable, it might seem irrelevant to debate the question whether terrorism can be morally justified. Groups like ISIS and Hamas have a perverted view of human flourishing, so there is no question of their terrorism being morally justified. However, as Christopher J. Finlay observes, ‘the chief point of analyzing the hypothesis that terrorism could be morally justifiable in some conceivable circumstances’ is ‘not to mandate or condone it in any historical cases, but to provide the theoretical equipment we need in order to be able to specify how such cases generally fall short of the necessary justifying conditions’.²⁶ One such justifying condition is that terrorism has morally worthy or at least acceptable objectives. While it is undoubtedly the case that many actual terrorists are deluded about the moral status of their objectives, I can think of no good reason to rule out the possibility of terrorism having morally acceptable or even worthy objectives. The real concern in such cases is not that terrorism has bad ends but that it is not an acceptable or appropriate means of pursuing even good ends. The question is why not.

III

One answer to this question is that terrorism is unlikely to succeed. Its terrible human costs are, in general, far more certain than any benevolent outcomes that are likely to be brought about by it, and it is a moral objection to terrorism that it imposes such costs—usually on others—when the likelihood of success is low.²⁷ In addition, there are almost always alternatives to terrorism. The implication is that terrorism, even for good ends, falls foul of (L) and (A) and therefore cannot be morally justified regardless of one’s view of its objectives. However, both (L) and (A) raise additional questions. For example, (L) says that terrorism is morally justified only if it is likely to achieve its objective but how likely does success need to be? Is there a reliable way of calculating the likelihood of success and how confident is anyone entitled to be about such calculations? What can we learn from the historical evidence?

Suppose that likelihood is understood as probability. On this understanding, (L) states that terrorism is morally justified only if there is a high probability that it will achieve its end. One issue that this principle brings to the fore is whether the probability of success is ever high enough for terrorism to be morally justified. A prior issue is that it is not possible to judge the relevant probabilities ‘with the precision needed for rational confidence’.²⁸ Does this matter? One might take the view that for a high probability of success to contribute to the justification of terrorism, there only has to be a high probability of success. It does not matter whether it is possible for terrorists or anyone else to know or have rational confidence that success is highly probable. On a different view, the mere fact that an act of terrorism has a high probability of achieving its end contributes little to the moral justification of terrorism if the probability of success is beyond the ken of the terrorist.²⁹

On the latter view, the difficulty of judging the relevant probabilities with the precision needed for rational confidence makes it extremely difficult for terrorism to be morally justified. Imagine a terrorist trying to work out the probability that a planned terrorist act will deliver the political change he desires. Suppose that for the act to be morally justified there must be at least a 50% probability of success. The difficulty is not just that this threshold is arbitrary but that there is no sensible way for the terrorist or anyone else to assign probabilities in this case. The general point has been well made by the distinguished economists John Kay and Mervyn King, who are critical of abuses of the notion of probability, defined as the quantitative expression of the likelihood of one of several possible outcomes.³⁰ Does (L) fall into what Kay and King refer to as ‘the modern trap of bogus quantification’?³¹

Kay and King note that the value of probability theory is well established in relation to games of chance and the analysis of data generated by a stationary process.³² Hence, it is possible to calculate the probability of exactly 500 heads when a fair coin is tossed a thousand times.³³ As soon as uncertainty is introduced, talk of the probability of a particular outcome can all too easily become a way of disguising our lack of knowledge. We live in a world in which there is too much that we do not know. Like emperors, explorers, and presidents, terrorists have to make decisions ‘without fully understanding either the situation they faced or the effects of their actions’.³⁴ This was bin Laden’s predicament. He had no way of knowing how America would react to 9/11. He had one set of expectations while other al-Qaeda leaders had a different set of expectations. None of the parties to the debate was in a position credibly to assign a probability to a particular outcome. Matters were further complicated by the difficulty of estimating the probability that the terrorists would manage to hijack four planes and fly them into their targets. However, on no reasonable view was there a high probability of tactical or strategic success, and so it is with most acts of terrorism. For most terrorists, the honest answer to the question ‘what is the probability of success?’ is ‘I don’t know’.

Does this matter? Suppose that for an act of terrorism to be morally justified, it is only necessary that there is a high probability of success, not that those contemplating the act can know or have rational confidence that success is highly probable. However, if nobody is able to calculate the probability of success, then nobody can determine whether an act of terrorism satisfies (L). Even after the event, it can be hard to tell whether any political change that follows an act of terrorism was caused by it. Furthermore, even if terrorism caused political change on this occasion, its success might have been a fluke. We will have more to say below about the significance of this in relation to (L). Terrorists who accept that the probability of success is low may feel some chance of success is better than no chance and that they are justified in taking a chance on terrorism even in cases where they are unlikely to be successful. However, when terrorists gamble like this they typically gamble with the lives of others. The inescapable reality is that taking another person’s life or maiming them in a political cause is more certain in most cases to produce pain than to generate the expected beneficent political change. This is both a moral and a practical objection to terrorism.

At least some of these difficulties are caused by the decision to interpret likelihood as probability. Kay and King argue against identifying likelihood with probability.³⁵ To use their example, a person with limited knowledge of American capitals might reason that it is likely that Philadelphia is the capital of Pennsylvania on the basis that the capital of a country or region is often its principal city. However, the statement ‘the probability that Philadelphia is the capital of Pennsylvania is 0.7’ is absurd.³⁶ When people are asked whether it is likely that such-and-such, ‘they do not reason probabilistically but interpret the question in the light of their broad contextual knowledge’.³⁷ The person who thinks it likely that Philadelphia is the capital of Pennsylvania goes wrong because in America the state capital is often not the principal city. The city of New York is not the capital of the state of New York, Los Angeles is not the capital of California, and Philadelphia is not the capital of Pennsylvania.

In cases where numerical probabilities cannot credibly be assigned to outcomes, people rely instead on narratives. Narratives are stories that humans tell to interpret complex situations. A reference narrative is ‘a story which is an expression of our realistic expectations’.³⁸ A credible narrative is one that is consistent with human experience and risk is ‘failure of a projected narrative, derived from realistic expectations, to unfold as envisaged’.³⁹ In ordering the 9/11 attacks, bin Laden relied on a reference narrative about how America would react to being attacked on such a massive scale. His narrative was flawed and failed to unfold as envisaged. The same goes for the Hamas leaders who ordered the 7 October attack on Israel. They presumably did not expect Israel to react in the way that it did, though it is not clear why not.

This points to a way of understanding (L) that exploits the idea of a reference narrative. On this account, when (L) speaks of terrorism being justified only if it is likely to achieve its objective, this is to be understood as claiming that an act of terrorism is morally justified only if there is a realistic reference narrative, based on broad contextual knowledge, that points to the likelihood of success. In this context, the likelihood of success is the likelihood that events will unfold as envisaged by the terrorist, leading to achievement of the terrorist’s objective. In too many cases, terrorist reference narratives lack credibility and the risk of events not unfolding as envisaged is high. However, this need not be so. The possibility of a credible and robust terrorist reference narrative cannot be definitively ruled out, and (L) is not a requirement that could not possibly be satisfied.

The contextual knowledge on which credible reference narratives are based includes historical knowledge. Research by Audrey Cronin into 450 terrorist campaigns shows that 87.1% of them achieved none of their strategic aims, 6.4% achieved a limited result, 2% achieved a substantial component of their aims, and only 4.4% has succeeded in the full achievement of their primary stated aim.⁴⁰ Terrorism is an instrumental business: people become involved in the terrorist process in order to achieve something else, but the historical record suggests that the results are, at best, patchy. A terrorist reference narrative that posits a likelihood of success flies in the face of history, and there is more than an element of wishful thinking in terrorists’ rosy reference narratives. If, as (L) suggests, terrorism is morally justified only if it is likely to succeed in strategic terms, then it is easy to conclude that most terrorism is not morally justified.

There are, however, exceptions. Terrorism sometimes works. To give one example, a good case can be made that the Irgun’s terrorism expedited the withdrawal of the British from Palestine and the establishment of the state of Israel, which was the Irgun’s primary goal.⁴¹ Thus, when a terrorist group’s reference narrative envisages the likelihood of success, an optimistic narrative might be justified on the basis that its situation is much more akin to that of the Irgun and other ‘successful’ terrorists groups than that of the many terrorists groups that, as Cronin shows, failed. This line of defence might be bogus but need not be. Perhaps there are contextual factors that point to the likelihood of success in a specific case regardless of whether terrorism has in general been successful. In cases where a terrorist group has good grounds to believe that its terrorism will buck the trend and is likely to be successful, it would have to be admitted that it might satisfy (L).

We have so far been discussing the moral justification of terrorism in the light of (L) as a prospective matter but there is also scope for retrospective justification. Suppose that a group that is planning a terrorist attack lacks a credible reference narrative that points to the likelihood of success. The likelihood of success is low, but the group is lucky and gets what it wants through terrorism. Would this be a case in which the group’s terrorism satisfies (L)? Terrorism, it might be argued, is in the results business and talk of the likelihood of success becomes irrelevant after an act of terrorism has been carried out and, however improbably, has delivered the desired result. As we have observed, it is always hard to be certain how much terrorism has contributed to political change but consider a case where one can be reasonably confident that terrorism did in fact make a positive difference, even though one would not have predicted in advance that it would do so. In such a case, what matters is surely not the prior likelihood of success but actual success.

This is effectively a case of what philosophers call moral luck.⁴² A drunk motorist who runs a red light and kills a pedestrian is seen as more blameworthy than one who runs the same red light in a drunken stupor but harms no one because no pedestrian was present. Yet it was sheer luck that nobody was harmed in the latter case. The interesting philosophical question is whether it is rational to regard the second motorist as less blameworthy than the first. The issue with (L) is whether, given two terrorist groups with equally weak reference narratives, a group that is successful because it gets lucky is less morally blameworthy for its terrorism—or more morally justified—than one whose terrorism ends in predictable failure. Intuitions are likely to vary. For readers who believe that moral standing can be a matter of luck, the lesson should be that (L) can be bypassed in some cases. Regardless of the prior probability of success, terrorism that hits its target has more going for it from a moral point of view than terrorism that does not, assuming that the target is morally defensible.

What about the case, mentioned above, in which the target or objective of terrorism is revenge? This is one of Richardson’s three Rs and her discussion implies that it is one target that terrorists cannot miss since they need rely only on themselves to get revenge. This cannot be quite right. Suppose that a terrorist group decides to take revenge against a given state by bombing its embassy in London. Through sheer incompetence, it bombs the wrong embassy. This would be a case where the group has tried but failed to take revenge. Still, Richardson’s basic point is correct: even in cases where they are unlikely to achieve the political change they seek, terrorist groups that seek revenge are likely to achieve it. Failure is not impossible, but revenge terrorism is ‘successful’ more often than not.

This has little bearing on the moral justification of terrorism if revenge is not a morally acceptable or worthy objective. Consider the following observation by the philosopher Robert C. Solomon:

Vengeance is the original passion for justice. The word “justice” in the Old Testament virtually always refers to revenge. In Kant and Hegel the word Gerechtigkeit refers to retribution, and throughout most of history the concept of justice has been far more concerned with the punishment of crimes and the balancing of wrongs than it has been with the fair distribution of goods and services. “Getting even” is and always has been one of the most basic metaphors of our moral vocabulary, and the frightening emotion of righteous, wrathful anger has been the emotional basis for justice just as much as benign compassion.⁴³

On the other side are philosophers who argue that revenge is a futile attempt to right previous wrongs and can never be morally justified. One might split the difference, as it were, and insist on a distinction between revenge and retribution. The latter has been defined as ‘punishment inflicted as deserved for a past wrong’.⁴⁴ To take revenge is ‘to retaliate a past wrong by making the offender suffer’,⁴⁵ and the offender’s suffering is a source of what has aptly been described as ‘vindictive satisfaction’.⁴⁶ However, unlike retribution, revenge sets no limit to the amount of suffering to which the offender is subjected.⁴⁷ Revenge can be ‘over the top’ and still be revenge. Shooting someone in response to a slight might be revenge but it is not retribution. Revenge is personal but ‘the agent of retribution needs no special tie to the victim of the wrong for which he exacts retribution’. Revenge involves ‘a particular emotional tone, pleasure in the suffering of another’. Retribution needs involve no such emotional tone. Revenge need not be general, but the imposer of retribution ‘is committed to (the existence of some) general principle (prima facie) mandating punishment in other similar circumstances’.⁴⁸

These distinctions might be used in defence of the idea that, unlike revenge, retribution can be a morally worthy objective, given the link between retribution and justice. Consider the following example: a Nazi battalion enters a town in Poland and massacres its entire Jewish population. In response, a group of partisans attack a truck carrying the soldiers responsible for the massacre. In this example of retributive terrorism, in which no judicial remedy is available to the partisans, it is difficult to argue conclusively that the soldiers did not deserve their fate or that justice was not served by an act of terrorism. Yet this does little to justify modern acts of so-called retributive terrorism. While there is little doubt that Osama bin Laden thought of the 9/11 attacks as retributive, most of the victims of the attacks were not responsible for the wrongs for which retribution was sought. The gleeful Hamas terrorists who rampaged through kibbutzim on 7 October 2023 no doubt derived vindictive satisfaction from their killing spree, but it would be preposterous to regard their actions as acts of retributive justice. The point of the example of the Polish massacre is to show how some acts of violent retribution might be morally defensible, not to suggest that recent acts of terrorism are morally defensible on the same basis.

Understood as moral objectives, retribution and revenge are intertwined with the more obviously political objectives of modern terrorist groups. For Hamas, for example, vengeful violence has a strategic rationale: to extract a high price from Israel and Israelis for its violence against Palestinians. Rage-driven revenge inflicts pain and makes Israelis feel less secure and more under threat. However, Hamas’ ultimate strategic objective is the destruction of Israel and the establishment of an independent Islamic Palestinian state covering the historic territory of Palestine.⁴⁹ This is the context for assessing its second-tier objectives: revenge and retribution. The abstract moral merits of retribution as merited punishment for past wrongs are null and void when the first-tier political objectives of those seeking retribution are, as in the case of Hamas, exterminatory or genocidal.

Turning next to (A), this can be dealt with more briefly than (L). (A) assumes that acts of terrorism are prima facie objectionable and should be avoided, if at all possible, on account of the human suffering they entail. To say that they should be avoided if possible is to say that they should not be carried out if there is a viable alternative to terrorism. A viable alternative would have to be a non-terroristic method of achieving the same objective or close to the same objective. The latter qualification is needed because it is possible that a non-violent alternative would deliver most of the political changes that the terrorist seeks but not quite all. In such a case, a relevant question is whether the respects in which the alternative fails to deliver exactly the same results as terrorism are politically significant. If they are, then terrorists may feel that there is no viable alternative to violence because they are unwilling to accept anything less than the complete package of political changes that they seek. More realistic terrorists will recognize the need for compromise, at least at the margins. They should also recognize the extent to which political changes that come about without violence are likely to be more widely accepted than ones that result from violence.

Another issue is timing. Suppose it turns out that there are non-violent routes to political change but it will take far longer to reach the terrorist’s objectives without violence than with it. The terrorist will read (A) as saying that terrorism is morally justified only if there are no realistic non-terroristic means of achieving its objective within the same timescale. Whether this is an acceptable amendment depends on the urgency of the problem to which terrorism is a response. If terrorism is a response to severe oppression or impending genocide, then patience may not be a virtue. However, if violence can achieve in, say, five years what non-violence can achieve in ten, and the current level of oppression is survivable, then the resort to violence may well be unjustified. In practice, such calculations regarding timescale are so speculative as to carry little weight. Of far greater weight is what the historical record reveals about the efficacy of terrorism in comparison to that of non-violent methods.

On this issue, pioneering research by Erica Chenoweth and Maria Stephan suggests that (A) is a condition for the moral justification of terrorism that terrorist campaigns are unlikely to satisfy.⁵⁰ I have already referred to the evidence that terrorism regularly fails to achieve its strategic aims. This needs to be set against the impressive evidence that ‘compared with the alternatives, nonviolent resistance is a stunningly successful method of creating change’.⁵¹ The umbrella term for non-violent resistance is civil resistance, defined as a ‘method of active conflict in which unarmed people use of a variety of coordinated, non-institutional methods – strikes, protests, demonstrations, boycotts, alternative institution-building, and many other tactics – to promote change without harming or threatening to harm an opponent’.⁵² It was civil resistance not terrorism that ended slavery in America. It was civil resistance not terrorism that persuaded the British to leave India. In South Africa, the ANC did carry out some terrorist acts but its campaign of civil resistance contributed far more to ending apartheid. On one definition of ‘success’, civil resistance has, according to Chenoweth, been twice as successful as violence at bringing about revolutionary political change.

Thus, when terrorists claim that their violence is morally justified because their cause is just and there is no viable alternative to violence, there is a simple and devastating response to their special pleading: not only is civil resistance an alternative to terrorism, but it is also for the most part a more effective, as well as a less lethal, alternative. If terrorists claim to have tried civil resistance and failed to make it work for them, one should ask whether they tried hard enough.⁵³ If they failed to get anywhere with civil resistance, was that because of a lack of public support for their cause? If so, and the reason that there is no alternative to terrorism is that not enough people want what the terrorists want, then that will be a reason to suspect that terrorism is not morally justified. From a moral point of view, it matters why there is no alternative to terrorism, not just that there is no alternative. This is not to say that it is impossible for terrorism to satisfy (A) in the right way, as it were. It is possible that there is no alternative to terrorism in pursuit of a just and urgent cause, and that the absence of a viable alternative is not just a reflection of a lack of public support. It would not have been unreasonable for the French Resistance to claim that when it came to fighting the Nazi occupation of their country there was no alternative to terrorism. However, such cases are few and far between and highly context specific. The ‘no alternative’ card is an easy one for terrorists to play but hard to play convincingly. Since it is far from certain in most cases that terrorism is the only viable option, it is far from certain that terrorism satisfies (A).

The destructiveness condition is no less challenging for terrorism. As with (A), it would be foolish to insist that it is impossible for terrorism to satisfy (D). However, the reality is that terrorism is often needlessly destructive. Terrorists take insufficient care to minimize the harms to innocent parties, and horrorists try to maximize such harms. Far too much terrorism is what I have called ‘maximal’ terrorism, and it is an interesting question why this is. To begin with, however, much more needs to be said about the notions of ‘minimal’ and ‘maximal’ terrorism. In particular, it needs to be recognized that what counts as minimal or maximal is objective-relative. An act of terrorism is morally justified only if it is a minimally destructive means of achieving its objective, and a minimally destructive means of achieving one objective might not be an effective means of achieving a quite different objective.

As I have noted, (D) runs into the following potential problem: suppose the terrorists’ objective is one that can only be attained by acts of extreme violence or horrorism. In that case, will we not be compelled to say that it is possible even for horrorism to fulfil the destructiveness condition? If the objective is genocide, then surely genocidal methods will count as minimally destructive relative to that objective. When I mentioned this concern previously, I rejected it on the grounds that (D) presupposes a morally acceptable objective and therefore excludes any attempt to justify extreme methods for attaining genocidal objectives. Still, could there not be cases where morally worthy objectives can only be attained by extreme violence? Did it not take extreme violence, albeit perpetrated by Allied soldiers, to attain the worthy objective of defeating the Nazis? One can imagine practitioners of horrorism arguing that it is sometimes necessary to terrorize the enemy even in pursuit of legitimate objectives, and that in such cases it is indeed possible for acts of horrorism to pass the destructiveness test.

I reject this suggestion. I doubt that there is any legitimate political objective that can only be attained by horrorism. Beheadings and such other attacks on the bodily integrity of victims are never minimal terrorism and are therefore never morally justified. There are moral red lines, and certain types of action are beyond the pale morally speaking regardless of their real or imagined practical utility. It might be wondered, however, whether there is any proper basis for distinguishing beyond the pale terrorism—such as that of Hamas on 7 October 2023—and garden variety terrorism? Some readers will protest that all terrorism is beyond the pale, morally speaking. After all, if the problem with horrorism is that it sets out not merely to kill its victims but to mutilate them, then how is this different from car bombing a busy street? In both cases, victims will be mutilated as well as killed, so what exactly is the difference morally speaking?

This is a reasonable question, which might be answered by highlighting the distinctive feelings of revulsion caused by ‘hands on’ acts of mutilation such as those carried out by ISIS and Hamas terrorists.⁵⁴ However, to the extent that there is no sharp dividing line between acts of horrorism and more familiar terrorist acts involving explosives, this only goes to show that both fall foul of (D). Terrorists make choices when they pick their weapons and select their targets, and they hardly ever choose minimal terrorism, as required by (D). They fail to satisfy (D) when they target people rather than property, civilians rather than combatants, and make no effort to minimize civilian casualties by giving adequate warnings. Even relative to the political objectives of ‘traditional’ terrorist groups such as the Provisional IRA, the methods used by such groups do not count as ‘minimal’.

This is partly the result of the fact that it is much easier in general for terrorists to attack ‘soft’ targets—pubs, commuter trains, shopping malls—than to attack the combatants to whom they are opposed. Callousness and a tendency to dehumanize victims also play a significant role. During its bombing campaign, the Provisional IRA gave warnings to the authorities, but its warnings were frequently too late or too vague to be actionable. As a result, large numbers of civilians died unnecessarily. Another factor in non-minimal terrorism is a desire for revenge or retribution. When terrorists like bin Laden convince themselves by spurious reasoning that their civilian victims ‘had it coming’, there is no incentive to ensure that minimally destructive means are used in pursuit of their political objectives. The methods employed are designed to maximize civilian casualties and, as such, can never be morally justified, regardless of whether the cause is just. (D) is a reminder that terrorists have options and that their actions cannot be morally justified when unnecessarily destructive methods are used.

I have so far considered four of the five elements of MODAL and my conclusion can be succinctly stated: the answer to the question whether terrorism can be morally justified depends on whether the (O), (D), (A), and (L) are necessary conditions for moral justification that it is possible for an act of terrorism to meet. A test of possibility is conceivability, so I have focused on whether it is conceivable that an act of terrorism satisfies the MODAL tests. I have concluded that this is indeed conceivable. In arguing for this conclusion, I have tried to avoid purely hypothetical or excessively far-fetched examples. I have tried to ground this discussion in an understanding of terrorism as it is and has been. However, this same understanding also reveals that while it is possible for terrorism to satisfy the MODAL test for moral justification, it is also difficult for terrorism to pass the test. The MODAL conditions I have identified are demanding, as they should be, and most real-world terrorism plainly fails the tests. There is one more matter to discuss, namely, the (M) element of MODAL. I shall now bring this paper to a close by considering the question of motive.

IV

According to the philosopher Michael Slote, ‘an act is morally acceptable if and only if it comes from good or virtuous motivation involving benevolence or caring’.⁵⁵ Where Slote speaks of the conditions for an act to be morally acceptable, (M) speaks of the conditions for an act of terrorism to be morally justified. I do not take the distinction between ‘justified’ and ‘acceptable’ to be significant. For Slote, having a good or virtuous motivation is necessary and sufficient for an act to be morally acceptable. (M) only identifies a necessary condition for an act of terrorism to be morally justified, namely, that the act in question has a morally sound or at least acceptable motive. There are several other conditions that would need to be satisfied for the act to be morally justified. What Slote calls good or virtuous motives are, in my terms, morally sound or acceptable motives. Finally, whereas (M) is unspecific about what counts as a morally sound motive, Slote identifies benevolence or caring about the well-being of others as the archetypal virtuous motive.

Before considering what a proponent of (M) should be prepared to regard as a morally sound or acceptable motive, there is a more fundamental question: does the rightness of an act of terrorism, or indeed any action, depend on the agent’s motive? Not according to J. S. Mill. He distinguishes between motive and intention and argues that the rightness or wrongness of an action depends on the agent’s intention but not their motive. Consider the case in which ‘a tyrant, when his enemy jumped into the sea to escape him, saved him from drowning simply in order that he might inflict upon him more exquisite tortures’.⁵⁶ In this case, what makes the tyrant’s action morally wrong is his intention, that is, what he wills to do, rather than his motive, ‘the feeling which makes him will so to do’.⁵⁷ If the tyrant willed to save the drowning man but not in order to torture him, then both his intention and his action would have been different. For Mill, the agent’s motive ‘makes a great difference in our moral estimation of the agent’ but not to our moral estimation of his action.

The debate between Mill’s view and the motivational approach to moral justifiability cannot be settled here, and there are many other ways of thinking about the role of motives in the moral justification of actions. My conception of a motive is somewhat different from Mill’s. As the philosopher Jonathan Dancy has insisted, a person’s motives are not their subjective states of mind—‘feelings’, as Mill calls them—but the considerations in the light of which they perform a given action and which they take to favour performing the action. A person’s motives are their motivating reasons, and it is not implausible that the reasons for which a person acts have a bearing on whether their action is morally justified. My concern here is not to offer a compelling philosophical defence of this idea. The question is a different one: assuming that (M) states a necessary condition for an act of terrorism to be morally justified, what would count as a morally sound or acceptable motive in this context?

Benevolence or caring are hardly to the point since these are not typically among the motives by which terrorism is motivated. Terrorists are much more likely to be motivated by a desire for revenge or retribution, and there are views of morality on which at least the latter is, or can be, a morally acceptable motive as long as the retribution is just and proportionate and sought for a genuine injustice. If a desire for retribution is a morally acceptable motive, and is what motivates some acts of terrorism, then it follows that some acts of terrorism have morally acceptable motives. Even if a desire for retribution is morally problematic, there is another way of making space for the idea that terrorism can have morally acceptable motives. For I have already argued that terrorism can have morally acceptable or admirable objectives. That being so, a person’s motive for committing a terrorist act would potentially be morally acceptable if the consideration in the light of which they acted and which they took to favour their action was that committing the act would contribute to the achievement of a morally admirable end. From the fact that a person’s motives are good, it does not follow that their actions are morally justified all things considered. From a moral point of view, their good motives might easily be trumped by the appalling consequences of their action or by other considerations. However, this is not to deny that it is possible for terrorist actors to have morally acceptable motives and therefore possible for (M) to be satisfied in a given case.

A serious concern about (M) is that it is too easily satisfied by acts of terrorism. Those who commit such acts always claim to have good motives. Their only motive, they say, is to do what is morally right or what morality demands. For example, retributive terrorists regard retribution as a moral demand. However, when terrorists have a mistaken view of what morality demands, their belief that their motives are good is also mistaken and makes them even more dangerous. They may become what one philosopher calls moral fanatics who are ‘wrongly convinced that morality requires something perverse, say, that they exterminate the members of some despised minority group’.⁵⁸ This has an important bearing on another important question: how and why do people become terrorists? The sobering truth is that a significant proportion of terrorists are motivated by the desire to do what is right even in cases where it seems completely obvious to us that they are wrong about what is right.

This is as much as needs to be said here about (M). The lesson of this discussion of this condition is that if it is indeed a genuine condition for terrorism to be morally justified then it is also one that it is possible for an act of terrorism to satisfy. As we have seen, the remaining four MODAL conditions are, in principle, also capable of being satisfied. Confronted by the latest terrorist outrage, readers may be appalled by the thesis that terrorism can ever be morally justified. I sympathize with the view that the worst outrages of groups like ISIS and Hamas can never be morally justified but it simply does not follow that terrorism can never be morally justified. Readers will have to make up their own minds whether any actual terrorist act has fully satisfied the conditions for moral justification or whether an act that passes the MODAL test might still not be morally justified. I offer these reflections in this paper as a practical framework for thinking about these important questions.⁵⁹